GraphDLG: Exploring Deep Leakage from Gradients in Federated Graph Learning

2026-01-27

AI Summary

This work addresses the privacy risk in federated graph learning where gradient leakage can lead to the reconstruction of both the original graph structure and node features. It presents the first efficient joint reconstruction approach for these two components. Through theoretical analysis, the authors demonstrate that the graph structure can recursively reveal node features, and they propose GraphDLG—a method that integrates closed-form recursive rules with gradient inversion techniques, leveraging either random or client-side local graphs as auxiliary information to enhance reconstruction fidelity. Experimental results show that GraphDLG improves node feature reconstruction by over 5.46% in MSE and graph structure recovery by over 25.04% in AUC, significantly outperforming existing methods and overcoming the limitations of traditional Deep Leakage from Gradients (DLG) when applied to graph-structured data.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Data Mining & Knowledge Management: Graph Mining, Social Network Analysis & Community
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Federated graph learning (FGL) has recently emerged as a promising privacy-preserving paradigm that enables distributed graph learning across multiple data owners. A critical privacy concern in federated learning is whether an adversary can recover raw data from shared gradients, a vulnerability known as deep leakage from gradients (DLG). However, most prior studies on the DLG problem focused on image or text data, and it remains an open question whether graphs can be effectively recovered, particularly when the graph structure and node features are uniquely entangled in GNNs. In this work, we first theoretically analyze the components in FGL and derive a crucial insight: once the graph structure is recovered, node features can be obtained through a closed-form recursive rule. Building on this analysis, we propose GraphDLG, a novel approach to recover raw training graphs from shared gradients in FGL, which can utilize randomly generated graphs or client-side training graphs as auxiliaries to enhance recovery. Extensive experiments demonstrate that GraphDLG outperforms existing solutions by successfully decoupling the graph structure and node features, achieving improvements of over 5.46% (by MSE) for node feature reconstruction and over 25.04% (by AUC) for graph structure reconstruction.

Problem

• Federated Graph Learning
• Deep Leakage from Gradients
• Graph Privacy
• Gradient Inversion
• Graph Reconstruction

Innovation

• GraphDLG
• Deep Leakage from Gradients
• Federated Graph Learning
• Gradient Inversion
• Graph Reconstruction